Will auto enrollment help in funding employee retirement? The jury is still out, with studies indicating that there’s more to the issue than meets the eye.

Previously we looked at scholarly research from the Center for Retirement Research at Boston College, which found that older workers who were auto enrolled in their retirement plans tended not to contribute as much to the plans as workers who had had to enroll themselves.

But research from Bank of America Merrill Lynch comes down heavily on the other side of the issue, saying that a secondary benefit of auto enrollment is that participants boost their contributions on their own: “In 2014, 46% of automatically enrolled participants voluntarily increased their contribution rates. In 2011, the figure was only 25%,” the research said.

There’s no question that more workers participate under an auto enrollment feature than under a plan that has no such provision. The BAML research found that, among the firm’s clients, plans with automatic enrollment had 32 percent higher average participation rates than plans that do not automatically enroll employees.

BAML also found that, year over year, there was an 8 percent increase in plans using auto enrollment, a 25 percent increase in plans that combined auto enrollment with auto increases and a 49 percent increase in plans that offer a standalone auto increase from 2013 to 2014.

The BAML research said, “Average contributions increase dramatically for participants who are automatically enrolled. Every year since 2009, contributions for automatically enrolled participants have increased.” It added that in 2010, 25 percent of auto-enrolled participants voluntarily increased their contributions, while in 2012–2013, 41 percent did so.

However, in a footnote, the paper also said that “dramatic increase” in average contributions “[i]ncludes both employer and employee contributions for participants who are automatically enrolled within that year.” So the increase is not solely due to participant actions.

The CRR study specifically looked at older workers, while the BAML research did not. When that’s considered, those on both sides of the argument could be correct, depending on the segment of the employee population being discussed. Regardless, it’s clear that additional data would be helpful in determining just how much auto enrollment benefits participants.